Jump to content

Template talk:Cases in the COVID-19 pandemic

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Reference proliferation

[edit]

This is currently showing as four separate references at Severe acute respiratory syndrome coronavirus 2#Epidemiology, even though all of the links are to JHU. Is there a way to alter the template so that the references can be grouped together under the same refname? Dekimasuよ! 14:28, 4 March 2020 (UTC)[reply]

Bump. Waddie96? Dekimasuよ! 11:55, 5 March 2020 (UTC)[reply]
Dekimasu I added a ref name and now they group, well spotted. comrade waddie96 (talk) 12:45, 5 March 2020 (UTC)[reply]

Can recovered cases be added

[edit]

It would help to add recovered cases to this. Regional case totals for Europe, North America, and Middle East might help as well, but that is more difficult to maintain. Bcharles (talk) 15:06, 10 March 2020 (UTC)[reply]

Bcharles, I have added recoveries to the template. Dekimasuよ! 15:37, 14 March 2020 (UTC)[reply]

Mainland China

[edit]

I am reverting an edit that seeks to subtract Hong Kong, Macau and Taiwan from the Mainland China total. Taiwan is reported separately from China by the source. Hong Kong and Macau are internationally recognized as part of China, and should not be designated differently from other provenances including Hainan. Bcharles (talk) 01:41, 24 March 2020 (UTC)[reply]

@Bcharles: I am trying to be consistent from place to place. These numbers are subtracted from "mainland China" at Template:2019–20 coronavirus pandemic data. United States Man (talk) 01:45, 24 March 2020 (UTC)[reply]
That template uses the name "China", which clearly includes Hong Kong and Macau. It should be changed on that template to be consistent with the internationally recognized national territory. Bcharles (talk) 02:34, 24 March 2020 (UTC)[reply]
@Bcharles: No it does not. It uses China (mainland), not "China" unqualified, or "China (PRC)", "People's Republic of China", or variants thereof. CaradhrasAiguo (leave language) 02:50, 24 March 2020 (UTC)[reply]
JHU was previously reporting them separately and had a specific category called "Mainland China". Sometime within the last few days, they have switched to simply "China". Originally, we were reporting mainland China because that's what the site had. Dekimasuよ! 03:56, 24 March 2020 (UTC)[reply]

Request for comment: Should we be including Hong Kong and Macau totals in the "Mainland China" section of the table

[edit]

Should we be including Hong Kong and Macau totals in the "Mainland China" section of the table?

In trying to keep consistent with Template:2019–20 coronavirus pandemic data and 2019–20 coronavirus pandemic in mainland China, I changed this total to subtract Hong Kong and Macau from the "mainland" total. This edit has been reverted. I feel that it is very important to be very consistent and having different totals for the same place is not consistent. The numbers from China come from this source: [1]. I would like to know what others think about this. United States Man (talk) 02:36, 24 March 2020 (UTC)[reply]

  • Mainland only (22 provinces, 5 autonomous regions, 4 municipalities): Mainland China by definition excludes HK, Macau, and Taiwan. The NHC source since 30 Jan has reported "31个省(自治区、直辖市)和新疆生产建设兵团报告新增确诊病例XXXX例,新增死亡病例XXX例(湖北XXX例),新增疑似病例XXXX例。", not 34 if those three regions were included. Every report has also included in the final line "累计收到港澳台地区通报确诊病例XXX例:香港特别行政区XXX例(出院XXX例,死亡X例),澳门特别行政区XX例(出院XX例),台湾地区XXX例(出院XX例,死亡X例)。", because HK and Macau have separate health governance structures. CaradhrasAiguo (leave language) 02:50, 24 March 2020 (UTC)[reply]
  • Mainland China should exclude HK, Macau, and Taiwan. The reason is that there are separate immigration borders and health systems. So the situation in these regions is quite different. Keep it consuistent and take off the 3 regions's numbers from the official .cn government stats. Graeme Bartlett (talk) 03:39, 24 March 2020 (UTC)[reply]
  • This template was originally set up to be simple (copy-and-paste from JHU), and there are only a few articles employing that field. Recently the number of changes required to keep it up to date is increasing rapidly: computing active cases, computing Chinese cases. Further, the China/Mainland China portion of the cases is now only a small percentage of the total confirmed cases. I would suggest splitting China from the template entirely. I object to using different sources for the data in this template because it makes all of the data internally inconsistent. Dekimasuよ! 04:02, 24 March 2020 (UTC)[reply]
    • The only thing is that with splitting China off this table, the places to which this table is transcluded will need to be adjusted to not having the China total as part of it anymore. However, that should be an easy adjustment. United States Man (talk) 04:51, 24 March 2020 (UTC)[reply]
  • The Government of the People's Republic of China is publishing figures for Hong Kong, Macau and Taiwan separately as addenda after the main report about the mainland (example). Secondary sources are reporting separately too. --MarioGom (talk) 09:09, 24 March 2020 (UTC)[reply]
  • The term mainland china can refer to the PRC (as opposed to the ROC, Taiwan), or to China excluding islands such as HK, Macau and Hainan. To be more clear i think the term China should be used to refer to the whole of the PRC. This is consistent with how other countries with outlying territories are handled (see US, UK, Nehterlands, etc.), and this is consistent with internationally recognized borders. I agree with Dekimasu above that separate numbers for China is not needed in this limited template. Bcharles (talk) 21:25, 24 March 2020 (UTC)[reply]

Semi-protected edit request on 28 March 2020

[edit]

With the current confirmed global cases at 621,636, It is the United States that has the largest amount of confirmed cases at 105,470, according to the Johns Hopkins COVID-19 map. Medistix (talk) 16:02, 28 March 2020 (UTC)[reply]

This info has been updated and the request is now outdated. United States Man (talk) 02:42, 30 March 2020 (UTC)[reply]

Anyway to automatically round numbers

[edit]

Anyone know? Doc James (talk · contribs · email) 02:40, 30 March 2020 (UTC)[reply]

Why do you want to round? United States Man (talk) 02:41, 30 March 2020 (UTC)[reply]
We want rounded versions in the text of the article suchas at 2019–20_coronavirus_pandemic
I tried but failed here User:Doc James/numbers Doc James (talk · contribs · email) 03:14, 30 March 2020 (UTC)[reply]
Oh I gotcha. So this can be transcluded with the rounded values to those pages using rounded numbers. United States Man (talk) 03:18, 30 March 2020 (UTC)[reply]
@Doc James: Look now and see if that's what you were looking for. United States Man (talk) 03:27, 30 March 2020 (UTC)[reply]
User:United States Man looks perfect :-) Doc James (talk · contribs · email) 03:33, 30 March 2020 (UTC)[reply]
How close are you wanting to round numbers? I figure we should round down to the nearest 100 or 500. At least don't round up for the place it says "at least". United States Man (talk) 03:35, 30 March 2020 (UTC)[reply]
I would say 500 now. Would be good to keep both JH and https://www.worldometers.info/coronavirus/ as sometimes one versus the other is higher. Doc James (talk · contribs · email) 04:10, 30 March 2020 (UTC)[reply]
  • I would prefer to have the rounding done on the page that is calling for the numbers, though having an option for rounding (like the option for an edit link) would be ideal. I am not familiar enough with the template syntax to add that. Here is what i found for rounding down the output:
{{formatnum: {{#expr: trunc ({{formatnum: {{Cases in 2019–20 coronavirus pandemic |confirmed ||ref=no}} |R}} / 1000) * 1000}}}},
which outputs: 676,609,000. This expression calls the number of confirmed cases without the reference link, removes the number formatting (commas), divides by 1,000, truncates the decimal portion, multiplies the result by 1,000 and reformats the number. One could then call the template without any parameters for the reference link.
It is easier to round the result, instead of rounding down. Use: {{formatnum: {{#expr: {{formatnum: {{Cases in 2019–20 coronavirus pandemic |confirmed ||ref=no}} |R}} round -3}}}} Bcharles (talk) 07:55, 31 March 2020 (UTC)[reply]
I sort of understand what you’re getting at but we have already went to all the trouble of setting it up this way and transcluded. To me, simple rounding is easy enough. I don’t believe long-winded setups such as you have there are any better than just transcluding a rounded number. Maybe because it’s still early, but I can’t wrap my head around this one. United States Man (talk) 14:12, 31 March 2020 (UTC)[reply]
@United States Man: thanks for helping with this! Ping me if need be in future. Best, comrade waddie96 (talk) 18:42, 31 March 2020 (UTC)[reply]
@Doc James and Bcharles: I did it using Rounddown (revision 948578800), I hope it's ok (it looks ok to me). Feelthhis (talk) 20:48, 1 April 2020 (UTC)[reply]

Source

[edit]

Sometimes https://www.worldometers.info/coronavirus/ is higher

Sometimes https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6 is higher

Each breaks from time to time. Thus might be good to use both. Doc James (talk · contribs · email) 04:11, 30 March 2020 (UTC)[reply]

Yeah I don't know what the big deal about this is. It has apparently been an issue for several days now. United States Man (talk) 04:15, 30 March 2020 (UTC)[reply]
Sometimes one is higher in one area and lower in another area, as well. For the articles it reduces consistency to have numbers often fall depending upon which source the latest editor has cited, or to have editors warring over which of the two sites to use. The numbers will all rise eventually. It is best to go with the same source over time (I prefer JHU, which was the source the template was originally based upon, had a Lancet article, and is cited by other MEDRS). To use both sources emphasizes the imprecision of the numbers, because they will always appear to be unverified by the second source.
When rounding as you have suggested here today, this is a bit less of an issue. But for articles where we are reporting numbers down to the final digit, there should be precision in what we are reporting. Indications that the trailing numbers are not significant figures don't strike me as a great thing to do when we are talking about how many people have died. If there is insistence on using two sources I will probably stop using the template and go back to adding the numbers manually in any articles I'm helping to maintain, as it was before the creation of the template this month. The emphasis should be on reporting numbers that were verified at a particular "as of" time, not on reporting the highest possible numbers. Dekimasuよ! 04:39, 30 March 2020 (UTC)[reply]
Fair enough. I just want to stay consistent from place to place. As high as the total is at this point, even if JMU briefly lags behind WOM it isn't going to make that much of a difference in the long run. United States Man (talk) 04:53, 30 March 2020 (UTC)[reply]

Yesterday JH was ahead. This am WOM is ahead by 5,000 cases. JH is more than 2,000 cases behind here[2] for the UK.

The thing about using both is it does show the imprecision. These numbers are imprecise so that is a plus IMO. Doc James (talk · contribs · email) 16:01, 30 March 2020 (UTC)[reply]

User:Dekimasu it is fascinating. JHs appear to pull ahead in the evening while WOM pull ahead during the day.
Did you notice that JH is partly based on WOM? One more reason maybe to use both. How quickly they update depends on who is sleeping I guess. Doc James (talk · contribs · email) 05:46, 31 March 2020 (UTC)[reply]
It's interesting because in the country articles, people are using all sorts of sources, like Tweets from the national Ministry of Health. There is no consistency across the ~180 country articles because editors often use the totals provided by local, official sources or in local papers/websites. Who knows, they might be more accurate for those locations than an international ticker that tries to compile global data. Liz Read! Talk! 05:54, 31 March 2020 (UTC)[reply]
I just wish JHU had data for deaths for the entire U.S., not broken down by state or local region. That is less useful for information you want at a glance. Liz Read! Talk! 05:58, 31 March 2020 (UTC)[reply]
I agree with Doc James That these numbers are imprecise, and are a moving target, so it is OK that there is variation among sources. It is healthy to have more than one source and to have some variation.
Liz, if you click on a country under the confirmed cases on JHU site, it will show the relevant numbers for that country. If you click on the country again it will go back to the global total. For the US, you can flip through country, state and county totals below the list of cases by country. Bcharles (talk) 06:56, 31 March 2020 (UTC)[reply]
The big thing is less reliability / accuracy but uptodate ness (they are all generally accurate at some point in time). Sometimes one is more uptodate. Sometimes the other is more uptodate. They are all going in the same direction. Doc James (talk · contribs · email) 15:34, 31 March 2020 (UTC)[reply]
Agree they are all going in the same direction, and for me JHU is a good source. If you'd like to add the above two as additional refs and use whichever has the highest number then I'm cool with that. comrade waddie96 (talk) 18:44, 31 March 2020 (UTC)[reply]
Thanks User:Waddie96 some do not want this template used if the higher numbers are not present and some do not want this template used if it uses anything other than JH. Not sure how to get consensus around this? Doc James (talk · contribs · email) 21:32, 31 March 2020 (UTC)[reply]

I was wondering how to grade/evaluate each source. My personal naive reasoning was like this.

"Center for Systems Science and Engineering (CSSE) at Johns Hopkins University (JHU). That feels authoritative." Therefore that should be the only one source, not only for consistency, but also for reliability.

"Worldometers.info? That does not feel authoritative at all. It feels like a random website." So I went to give that website a quick look and found out that there is a source for every single daily case all around the world. Obviously I did not check every single source, but that looked reliable.

So I don't know how I feel anymore as to which source is better. I also wondered who is responsible for worldometer and how they keep it, because it takes a hell of a collective job to keep such enormous volume of data organized and minutely updated.

So, in the end, now I prefer the CSSE@JHU over Worldometer less because of the authoritative and academic background (therefore methodical and rigorous) but more because I have no idea about who's behind Worldometer.

TL;DR Does anyone know who's behind worldometer and how they keep their numbers?

The only reason I see for keeping both references is if they're truly independent between themselves (CSSE@JHU and Worldometer) and both equally reliable, otherwise I think we should use only one to keep the numbers in wikipedia (referencing both it's ok). It's too weird to have the numbers decrease by the hours even if they'll eventually increase by the days. Numbers increasing and decreasing could be edit warring ("this source is better", "no, this source is better", etc). Let's at least try to agree with a single source to keep the numbers. My vote is for CSSE@JHU should no evidence to the contrary be given.

Feelthhis (talk) 23:31, 31 March 2020 (UTC)[reply]

FWIW, there is no "up and down" of the totals. We never decrease, only increase, so there isn't any supposed edit-warring or source-warring going on here. United States Man (talk) 23:34, 31 March 2020 (UTC)[reply]
JH just pulled ahead of WOM as it appears to every evening US time. Which is most uptodate likely simple depends on who is awake. Doc James (talk · contribs · email) 23:44, 31 March 2020 (UTC)[reply]
We never decrease, only increase If that's always the case that'd be fine, but let's be honest you can't speak for all wikipedia editors and eventually the roller coaster will kick-in (who knows how many times it went up and down already). Feelthhis (talk) 00:02, 1 April 2020 (UTC)[reply]
Well you can look in the edit history. I don't know anyone who has edited this template who would subtract numbers from the total, but carry on... United States Man (talk) 00:57, 1 April 2020 (UTC)[reply]
For consistency, simplicity and credibility, i think that use of one source, JHU, for this template makes sense. Use of other sources on other pages is fine, even when the numbers vary. That is why a source is linked, to show where the numbers came from. There is no one true number, and the larger number is not necessarily the most accurate. Bcharles (talk) 10:58, 1 April 2020 (UTC)[reply]
@Bcharles: if you excuse me I flat out copy pasta'ed your first sentence inside my revision. Cheers Feelthhis (talk) 20:44, 1 April 2020 (UTC)[reply]
The links below help a bit in explaining their own methods:
https://www.worldometers.info/about/
https://systems.jhu.edu/research/public-health/ncov/
Worldometers link seems like a self-promoting advertisement. JHU link seems more like a proper informative description:

Before manually updating the dashboard, we confirm the case numbers using regional and local health departments [...]All manual updates (outside mainland China) are coordinated by a team at JHU.

That's a fair justification as to why JHU lags behind. There are method, scrutiny and validation behind CSSE@JHU numbers (that was expected, as I said in my first comment).
Also, JHU does use Worldometer as a source (as noted by Doc James), so they're not independent from each other. The Worldometer data is being used by CSSE@JHU, after all. There's no need to reference that again in wikipedia if we opt to use only CSSE@JHU data. In the end, CSSE@JHU being the single source will encompass Worldometers data (as opposed to exclude it).
Considering all of the above and everything else in this discussion, I took the plunge and committed a revision in which I asked to only use CSSE@JHU numbers and removed the Worldometers reference (no need to keep it, as explained above). Feelthhis (talk) 20:40, 1 April 2020 (UTC)[reply]
I don’t necessarily agree that we should solely use one source. People aren’t idiots and will understand that the two numbers can differ from two different agencies or whatnot. Still though, nice to effectively wake up to a “consensus” being delivered within a few hours.
For people viewing this site on mobile, the Johns Hopkins website is a complete shitshow, whereas the WOM site loads perfectly and shows purely the statistics that the average reader would be looking for. Perhaps you guys should think of the “bigger picture”, rather than circlejerking around a singular source, especially when the page overall has roughly 500 of the damn things?
Also, it breaks reference number 5 on the main article. Perhaps this rather kneejerk reaction to delete it should be undone? - J man708 (talk) 22:46, 1 April 2020 (UTC)[reply]
I'm not very inclined to listen to much of what you had to say considering to uncivil language you just used. I haven't seen you here for days on end helping update these templates and build this one up. United States Man (talk) 22:50, 1 April 2020 (UTC)[reply]
We are asking to use singular source only for this template (not in the Main article). Reference in the Main article is fixed btw (beaten again :) Feelthhis (talk) 22:59, 1 April 2020 (UTC)[reply]
That’s not how Wikipedia works. Regardless of whether you’ve seen me a million times or not at all, my points are valid. I’m annoyed because this CONSTANTLY happens, when you are in a place like Australia or New Zealand and a conversation goes on over the space of what is night for you and then changes are effectively brought forth without your ability to retort to the Americans and Europeans who brang in said changes. It happens far too often. - J man708 (talk) 23:03, 1 April 2020 (UTC)[reply]
Well you're the only one of the 7 or 8 around here that hasn't agreed, and you didn't discuss anything before now so we didn't know we had to wait on you. Even with your arguments, the comments of all the others here would indicate we've reached a consensus. United States Man (talk) 23:07, 1 April 2020 (UTC)[reply]
Perhaps you should read WP:Vote? Once again, I woke up to this. I’ve not been watching this talk page actively. All I’m saying is that the JHU site is absolutely fucking terrible for anyone who reads this on mobile, which is majority of the public, I daresay. I’m not vouching that the website is bad, just that it formats atrociously. WOM at least works for people to see the numbers immediately. If we could organise a way to get the mobile site version to auto-load for people who view on mobile, then that would be fantastic, but I don’t think it’s possible. - J man708 (talk) 23:17, 1 April 2020 (UTC)[reply]
The ultimate goal is to make a good encyclopedia-like article and the CSSE@JHU provides the numbers to build such quality article. What does an external link appearance has to do with it? Well, you could add a section "Pretty external link for mobile view" and just insert the external link there then. Feelthhis (talk) 23:36, 1 April 2020 (UTC)[reply]
Can I ask, have you tried loading it on a phone so far? It’s really not user-friendly. I wouldn’t normally care (like having other less user-friendly sources such as books listed), but this is an extremely high traffic page, which is why I don’t think that JHU source is the perfect thing for this article. I get that the WOM source isn’t as accurate as the JHU, which makes this tricky. - J man708 (talk) 23:46, 1 April 2020 (UTC)[reply]

I've loaded it on a phone numerous times. It works just fine in landscape view... United States Man (talk) 23:51, 1 April 2020 (UTC)[reply]

It does work a lot better in landscape, but it’s still not great. There’s obviously no way we can have this auto-load in landscape for mobile users, but is there anyway we could load the mobile site automatically for people on phones? That would seem to be a pretty good compromise if possible. - J man708 (talk) 00:41, 2 April 2020 (UTC)[reply]

Straw poll

[edit]

Please list if you think we should use just one or both sources.

I think the WHO numbers lag dreadfully behind from what I've seen and noticed around here on Wikipedia with others mentioning it. So maybe that isn't the best to use. United States Man (talk) 00:41, 2 April 2020 (UTC)[reply]
Lame. How about the CDC or even the UN? Do they have the same issues? - J man708 (talk) 00:52, 2 April 2020 (UTC)[reply]
CDC is also very slow. United States Man (talk) 01:22, 2 April 2020 (UTC)[reply]
CDC doesn't even report state-level information, they admit their information is not complete. They don't even publish on weekends.Rwat128 (talk) 15:59, 10 April 2020 (UTC)[reply]
JHU sources include WOM and 1point3acres, so there is no need for redundant information. Rwat128 (talk) 15:59, 10 April 2020 (UTC)[reply]
  • JHU. JHU is a reliable source including under more stringent requirements, checks and incorporates the data from the other source (this is a reason to use JHU, not to use both, since it makes JHU more comprehensive), was the original source being used, and the point is not to post the largest numbers at any given time, but to present a consistent and verifiable set of numbers. JHU has been fulfilling this requirement for a few months now, and I do not believe it is necessary to change the template. Using two sets of numbers leads to changing them back and forth between sites with little or no benefit to the readership. Dekimasuよ! 01:18, 2 April 2020 (UTC)[reply]
  • Both- For the same reasons given by DocJames, and the fact that the Worldometer site is much more user-friendly. I believe the average Wikipedia user would benefit from having the opportunity to choose from either of the two. We don't have to have a 100% perfectly updated source, but on occasions, the jhu site has been 10-15k out and for time periods, imo, longer than desired. MattSucci (talk) 01:57, 2 April 2020 (UTC)[reply]
  • Seeing as everyone who edits here is aware that the numbers are, in the long run, practically the same, then surely it's ok to take the most up to date one. Maybe a small footnote could be added explaining this. MattSucci (talk) 04:50, 2 April 2020 (UTC)[reply]
  • Having a higher number in and of itself does not necessarily mean a source is more up-to-date, though. And as I pointed out in previous sections, sometimes one source has a higher number of recoveries and the other has a higher number of cases at the same time, etc. Dekimasuよ! 08:02, 2 April 2020 (UTC)[reply]
JHU sources include WOM and 1point3acres, so there is no need for redundant information. Rwat128 (talk) 15:59, 10 April 2020 (UTC)[reply]
  • JHU - "For consistency, simplicity and credibility, i think that use of one source, JHU, for this template makes sense"; to quote myself. Bcharles (talk) 05:24, 2 April 2020 (UTC)[reply]
  • JHU: both inflate totals because of territory double-counting as well as improper inclusion of suspected cases (instead of confirmed cases) in some countries, but let's keep it simple and stick to one. --MarioGom (talk) 21:40, 5 April 2020 (UTC)[reply]
  • JHU First off, JHU's map is actually published in a peer-reviewed journal called Lancet[1] (one of the most prestigious journals in science) whereas WorldOMeter.Info sounds like a fake website from the early 2000's peddling gimmicky information. Second, WorldOMeter is a FOR-PROFIT entity that sells live-counters for $35 subscriptions, should not give them advertisment space on Wikipedia. THIRD, WorldOMeter asks for User submissions for new cases (crowd-sourcing information), and this method has not been independently verified. FOURTH, Google now prominently features Wikipedia's Main Table on it's google searches (see: "Coronavirus cases" in Google), so we need a authoritation, credible, and reputable source. FOURTH, JHU's page already cites WorldOMeter and 1point3acres as sources. FINALLY, JHU is the most commonly cited tracker by governments, academics, corporate media, etc... in the world. Rwat128 (talk) 15:47, 10 April 2020 (UTC)[reply]

References

  1. ^ Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis; published online Feb 19. https://linkinghub.elsevier.com/retrieve/pii/S1473309920301201

Let's work with significant digits instead of "nearest 1000"

[edit]

For example, if we work with 3 significant digits, then we have:

For 857,579 cases, 857000 rounded down (no difference here). For 49187 deaths, 49100 rounded down (instead of 49000).

I think this way is more correct without harming the "broad figure" philosophy. What do you think? Feelthhis (talk) 23:52, 31 March 2020 (UTC)[reply]

  • In my opinion, cases, recoveries and deaths should be rounded down to the nearest 500, as these numbers are becoming so large, the need for preciseness is simply unnecessary. MattSucci (talk) 02:46, 1 April 2020 (UTC)[reply]
  • I agree that three significant digits is easy for readers to grasp and is consistent among varied numbers. So cases and recoveries would be rounded down to thousands, and deaths down to hundreds, though soon cases would be rounded down to ten thousands. Bcharles (talk) 10:46, 1 April 2020 (UTC)[reply]
  • I now agree too that cases and recoveries should be rounded down to thousands and deaths for the time being down to hundreds. As user Bcharles says, soon cases will probably have to be rounded down to 10,000s. MattSucci (talk) 11:04, 1 April 2020 (UTC)[reply]

Template with values from Wikidata

[edit]

I created (in case you think it is interesting) a totally different version of this template ca:Plantilla:Cases in 2019–20 coronavirus pandemic. Using the values entered in Wikidata and a module. This template is also used in the second paragraph of the corresponding article (ca:COVID-19). The module reads and checks the last values of Wikidata for deaths and cases. --Jmarchn (talk) 16:34, 2 April 2020 (UTC)[reply]

Jmarchn: Thanks! Good to know. I think Wikidata fits perfectly for this case, although last time I checked, the enwiki community was not extremely fond of Wikidata usage in the default namespace. --MarioGom (talk) 21:55, 5 April 2020 (UTC)[reply]
Thanks comrade waddie96 (talk) 14:50, 7 April 2020 (UTC)[reply]

Rounded numbers: Millions and Thousands

[edit]

@Ythlev: added millions, I added thousands. Best for readability in my opinion. Everyone ok with it? Feelthhis (talk) 14:49, 3 April 2020 (UTC)[reply]

That's gonna be a no from me. I have reverted the changes since no prior discussion was offered. The table looks good as it is. We have been changing the format here too much lately. The information presented now is clear and concise, so no need for further changes. United States Man (talk) 14:54, 3 April 2020 (UTC)[reply]
To be honest, no. I prefer numerals, but that's just me. MattSucci (talk) 14:57, 3 April 2020 (UTC)[reply]
I completely disagree with this. Displaying 1,030,000 inside paragraphs is non sense. 1.03 million is much better and easier to read and I thought this was common sense. I think we should vote. Feelthhis (talk) 14:58, 3 April 2020 (UTC)[reply]
I don't think it is "nonsense". It works fine either way, so I will support the current way. And I actually think what you did for the "thousand" makes it look worse. People, or least me, don't generally write "55 thousand" instead of "55,000". That just looks silly to me. United States Man (talk) 15:01, 3 April 2020 (UTC)[reply]
I bet the typography experts would agree with me on this... It has its place in tables or other formats, but on running text 102,010,347,431.343 or 0.0000001234 is non sense. Feelthhis (talk) 15:06, 3 April 2020 (UTC)[reply]
Yes but we don't have numbers that high, do we? :) United States Man (talk) 15:07, 3 April 2020 (UTC)[reply]
How about just for the millions (*edit: and hundreds of thousands), can we settle with that? Deal? [shake hands, don't leave me hanging] Feelthhis (talk) 15:09, 3 April 2020 (UTC)[reply]
No, sorry. "5 hundred thousand" is atrocious. United States Man (talk) 15:12, 3 April 2020 (UTC)[reply]
No, I mean 500 thousand. Ok, just for the millions. Feelthhis (talk) 15:13, 3 April 2020 (UTC)[reply]
(ec) If all this keeps up, I'll just start supporting the notion that we quit messing with this template like @Dekimasu: said a few days ago. I am tired of changing the format everyday. United States Man (talk) 15:11, 3 April 2020 (UTC)[reply]
This is not messing, this is refining and improving this template! Feelthhis (talk) 15:14, 3 April 2020 (UTC)[reply]
Have your consensus. But this is wrong. 1.02 million is better than 102 followed by 0000000000000000 inside running text. I see I'm alone in this but I hope others come to support me. You guys did it congrats, I'm officially frustrated >:( Enjoy it. Feelthhis (talk) 15:30, 3 April 2020 (UTC)[reply]

Poll: 1,020,000 or 1.02 million

[edit]

Please list your vote, each example in paragraphs below.

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt 102,000,000 ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

or

Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor incididunt 1.02 million ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

Option for thousands

[edit]

Could use them for this video script Wikipedia:VideoWiki/Coronavirus_disease_2019 otherwise it says "0" "0" "0". Doc James (talk · contribs · email) 01:59, 5 April 2020 (UTC)[reply]

User:Primefac I want an output that will produce "69 thousand" explicitly for the video script. Otherwise the text to speech engine does not work possible. Would you be able to add that option? Doc James (talk · contribs · email) 00:13, 6 April 2020 (UTC)[reply]
Never mind figured out how to add it. Doc James (talk · contribs · email) 00:25, 6 April 2020 (UTC)[reply]
@Doc James: Good job, if you need 3sf for your thousands I can do it tomorrow if no one else already does it (ps: tomorrow rounded active is very likely to reach millions, so I updated it and it's ready for it). Feelthhis (talk) 02:16, 6 April 2020 (UTC)[reply]
I guess I'll watch this main page, since I keep getting pinged... ;-)
I added in options for the active cases to have words for both thousands and millions since it's not at 1mil yet. Primefac (talk) 02:42, 6 April 2020 (UTC)[reply]
Perfect. Thanks everyone :-) Doc James (talk · contribs · email) 03:18, 6 April 2020 (UTC)[reply]

Only having to put in the number once

[edit]

Tried this https://en.wikipedia.org/w/index.php?title=Template:Cases_in_2019%E2%80%9320_coronavirus_pandemic&type=revision&diff=948894331&oldid=948894073&diffmode=source

But it did not work. Anyone else have other ideas? I guess we could use one more template? Doc James (talk · contribs · email) 16:39, 3 April 2020 (UTC)[reply]

Yes that is perfect :-) Doc James (talk · contribs · email) 01:41, 4 April 2020 (UTC)[reply]
Thank you for this comrade waddie96 (talk) 14:44, 7 April 2020 (UTC)[reply]

Proposed merge

[edit]

I think we should merge this template page with the page Template:Territories affected by the 2019-20 coronavirus pandemic in order to have a central template for basic coronavirus data. Sam1370 (talk) 09:42, 8 April 2020 (UTC)[reply]

Sure. But we should base that on the JH numbers. Doc James (talk · contribs · email) 15:03, 10 April 2020 (UTC)[reply]
 Done. Territories are now added to this template. {{Territories affected by the 2019-20 coronavirus pandemic}} is now just a shorthand. --MarioGom (talk) 20:45, 18 April 2020 (UTC)[reply]
@MarioGom: Thanks! But you forgot to noinclude the reference on the original template (the cases template auto cites) so it messed up a couple of pages. Fixed it, thanks again for merging. sam1370 (talk) 23:32, 18 April 2020 (UTC)[reply]

Ratio: should we really provide it?

[edit]

I'm coming from Talk:2019–20 coronavirus pandemic by country and territory § Problematic statements. Should we really provide case fatality rate from this template? It is being used for misleading claims. At this point, using cumulative cases for case fatality rate, when most cases are still active, is problematic. I know generalist press is doing this, but WP:MEDRS are way more cautious. See, for example, Estimating case fatality rates of COVID-19. --MarioGom (talk) 17:58, 9 April 2020 (UTC)[reply]

The ratio yielded by this template is not a case fatality rate and hopefully isn't being described as one. I'm not sure whether there are other ways in which the ratio given here might be useful. Dekimasuよ! 14:36, 10 April 2020 (UTC)[reply]
Dekimasu: That is exactly the way it was used at 2019–20 coronavirus pandemic by country and territory: The crude (preliminary) case fatality rate [...] ([4]). --MarioGom (talk) 14:41, 10 April 2020 (UTC)[reply]
I just removed this value's use in the infobox of Coronavirus disease 2019 for this reason. There's no reason, using the JHU data, that we can't correctly calculate CFR as (deaths / (deaths + recovered)). Whether or not it's valid to include in articles is still up for debate, but at least then we're discussing the right ratio. Bakkster Man (talk) 18:12, 22 September 2020 (UTC)[reply]
My two points from Talk:Coronavirus disease 2019 still apply to the correctly calculated CFR: 1. What gets counted as a 'case' is different depending on the territory and testing rates and 2. The CFR and IFR need to be clearly distinguished, since a lot of people (at least anecdotally from my personal experiences) treat CFR like IFR or vice-versa. I would urge caution using it in articles. It needs to be clear what it means and it shouldn't be given more weight than the quality of the underlying data warrants. Ashorocetus (talk | contribs) 18:35, 22 September 2020 (UTC)[reply]

Argcis number decline

[edit]

Why are the numbers going down? Is this due to massive miscalculation by JHU? Or a technical error on the website? Valoem talk contrib 18:37, 9 April 2020 (UTC)[reply]

Valoem: there is an incorrect inflation of cases at JHU and Worldometer in the order of tens of thousands (summary). They partially corrected the largest source of incorrect case counts (France) yesterday, but today they introduced the error again. I think we'll see some more swings in numbers as these errors are corrected. --MarioGom (talk) 12:30, 10 April 2020 (UTC)[reply]
User:MarioGom thanks for looking into this. Doc James (talk · contribs · email) 14:59, 10 April 2020 (UTC)[reply]
Update: The CSSE has partially acknowledged the issue ([5]) and it seems they fixed it partially ([6]). However, their figure for France is still double-counting cases and inflated by roughly 30k cases. Hopefully they can fix it completely soon. --MarioGom (talk) 14:11, 15 April 2020 (UTC)[reply]

There is very important background information such as "Map Information" and "Map FAQ" on the https://coronavirus.jhu.edu/map.html that is not available on the ArcGIS embedded URL. I suggest using the citation link to https://coronavirus.jhu.edu/map.html so people can see "Map Information" and "Map FAQ" instead of the ArcGIS link. Rwat128 (talk) 21:33, 11 April 2020 (UTC)[reply]

I see your point, but [7] is linked at the bottom of the ArcGIS page, and the FAQ is also linked directly there in the second line of text. I am not sure this is an improvement since it makes the map (what we are actually citing, for its data) significantly smaller on my device. Dekimasuよ! 02:55, 12 April 2020 (UTC)[reply]
Yea, but for the average-lay person, it's really hard to navigate the bottom section, they might never notice that information since it's a tiny box. https://coronavirus.jhu.edu/map.html just been updated this morning. The new section adds "Critical Trends", which tracks Critical Data with up-to-date visuals that give context to the data collected on JHU's COVID-19 map, such as new cases, mortality, cumulative cases ,and animated maps (VERY COOL). I think for the average lay-person, this will give them more visualization choices, so they don't have to keep going to WorldOMeter or 1point3acres for other visualization.Rwat128 (talk) 01:35, 13 April 2020 (UTC)[reply]
Update, they also added a brand new section for "US Maps." on https://coronavirus.jhu.edu/map.html which I think it's really helpful visualization for average person.Rwat128 (talk) 18:18, 13 April 2020 (UTC)[reply]

Why a US University over the WHO?

[edit]

Hi! I'm just wondering why we're using John Hopkins University statistics instead of the WHO's official data? While undeniably John Hopkins is a reliable source, the WHO is the international organisation for health and is, what I would imagine, the most reliable and valid source for statistics around COVID-19. On account of the international nature of WHO, against the fact that John Hopkins is mononational, I'd suggest that we make the change to use the WHO's statistics (https://covid19.who.int/). I think it's also worth mentioning that the John Hopkin's site is from the Systems Science and Engineering department of John Hopkins, not a health or medicine department. The site also references local media reports for figures, which in turn have referenced Facebook posts. It also is primarily US-relevant data (they only list testing numbers for US states etc). ItsPugle (talk) 08:30, 18 April 2020 (UTC)[reply]

ItsPugle, there are a few reasons, and some issues as well. Some background:
  • JHU data is used because (1) it is more up-to-date and (2) includes data about recoveries.
Regarding accuracy, the good parts:
  • JHU is consistent with WHO Situation Reports for most (not all) countries. With consistent I mean that at the daily cut-off for each country, cases and deaths usually match what WHO Situation Reports publish a few hours later. That is because WHO reports confirmed cases as reported by each country authority, and JHU publishes these figures first.
  • Note that most of these "Facebook posts" are official publications from official communication channels of the Government or Ministry of Health. In some countries, that is the primary public communication channel. In some cases, these official communications on Facebook include scanned documents signed by the relevant authority.
The bad parts:
  • JHU CSSE, per their own definition, includes both confirmed and probable cases. They do this for a few countries only (e.g. France, Canada, New Zealand). This is not in line with WHO Situation Report and it is the main source of divergences. Note that probable cases are not included for all countries where data is available. It is included only in a handful of them without a clear criteria. France is the most prominent example because the additional counted cases are almost 40k by now.
  • JHU CSSE has other minor issues, but they are negligible for world totals. See Wikipedia:WikiProject COVID-19/Case Count Task Force § Case double-counting.
  • Many countries do not have any stable source for information about recoveries. For example, the United Kingdom. JHU CSSE reports very low values for these, based on old data. Anyway, I think we can assume that recoveries, as reported by JHU CSSE is always a lower bound of the real value.
Personally, I think that using JHU CSSE for world totals, including recoveries, is a reasonable compromise. However, given the known issues with accuracy, we should only use the rounded values from the template, cited as approximately 2,250,000 cases, instead of 2,256,844 cases. --MarioGom (talk) 10:39, 18 April 2020 (UTC)[reply]
@MarioGom: Fair enough. I guess. I still have my hesitations on the points you raised (and as far as I know of, the country I'm currently residing doesn't report their figures to any agency other than the WHO), but I think your idea of using WHO data in another template as an alternative for JHU is a great idea. ItsPugle (talk) 01:33, 19 April 2020 (UTC)[reply]
ItsPugle, it probably does report figures publicly before reporting to WHO. See {{2019–20 coronavirus pandemic data}} where we have individual sources for every country. --MarioGom (talk) 08:14, 19 April 2020 (UTC)[reply]
@MarioGom: I don't think so? Each Australian state reports their data to the national Health Department and the National Coronavirus Co-ordination Commission as cases change, and at 3pm each day the changes/updates are reported to the WHO and made public on the Health website (I've also just updated that template with the latest update from Health). Nonetheless, let's see how your idea below goes! Thanks, ItsPugle (talk) 08:33, 19 April 2020 (UTC)[reply]

New template for latest cases data from WHO

[edit]

Following up #Why a US University over the WHO? as well as quite a few previous discussions, I think we should create a second template (maybe {{WHO cases in 2019–20 coronavirus pandemic}}, or {{Cases in 2019–20 coronavirus pandemic/WHO}} that is similar to this one, but including only data from the latest WHO Situation Report. It would include "total confirmed cases", "total new confirmed cases", "total deaths", "total new deaths", "report number" and "report date". It would need a single update per day. It could replace the usage of the current template at Coronavirus, Severe acute respiratory syndrome coronavirus 2, Coronavirus disease, etc. In the infobox at 2019–20 coronavirus pandemic we could use both templates with WP:INTEXT attribution. In {{2019–20 coronavirus pandemic data}} we would continue using JHU CSSE for totals, since they are closer to the sum of the table and the nature of "rolling reports". What do you think? --MarioGom (talk) 11:06, 18 April 2020 (UTC)[reply]

I think this is a very good idea for having a dual representation of JHU's "rolling reports" and the official international statistics from the WHO. I would probably move to rename {{Cases in 2019-20 coronavirus pandemic}} to {{Rolling cases in 2019-20 coronavirus pandemic}}, and in addition have the WHO data under {{Official cases in 2019-20 coronavirus pandemic}}, but that's simply a matter of preference and can be sorted out at a later stage. ItsPugle (talk) 01:37, 19 April 2020 (UTC)[reply]
Okay, I've gone ahead and created the WHO version of this template. I literally had no idea what I was doing (hence why the template looks suspiciously similar in literally every aspect to this template), so I would appreciate it getting a once-over by someone who knows how this all works properly! ItsPugle (talk) 07:28, 20 April 2020 (UTC)[reply]
Update on this: I've done the best I can, but I don't know how to remove the commas from the data that's included (because the WHO dashboard doesn't have recovered or territories) so that's breaking it. It also has a grey background on the template page, which tells me that there's a code block or something funky (I couldn't figure it out), so I would appreciate some help there as well if anyone knows what's going on. ItsPugle (talk) 08:10, 20 April 2020 (UTC)[reply]

Suggestion to remove the colon in the 24h timestamp

[edit]

Hi! I'd suggest that we move to align the 24h time syntax to the format as described in ISO 8601. The international standard (as in ISO 8601) puts hhmm as the basic form of a 24h time representation, with hh:mm:ss (and in this hh:mm) as an extended form should it be needed. Generally in Commonwealth English aswell, at least in all my experience, the use of a colon is used to signal a 12h timestamp not separate the hours from minutes, as the first two digits of a 4-digit 24h timestamp are always the hour and the last two are minutes. In my readings of and contributions to Wikipedia, this has never been an issue or any sort of consideration, especially given the context of the change, but if there is a pre-existing consensus that would formalise the current syntax, please share! Thanks, ItsPugle (talk) 07:28, 19 April 2020 (UTC)[reply]

MOS:TIME makes clear that a colon separated time(stamp) should be used. Fred Gandt · talk · contribs 19:14, 19 April 2020 (UTC)[reply]
Ah, okay! As I said, I wasn't aware of any existing consensus about this matter, but thanks for sharing that :) ItsPugle (talk) 23:22, 19 April 2020 (UTC)[reply]

Proposed renaming to COVID-19 pandemic

[edit]

Hi there, since the main talkpage (COVID-19 pandemic) actually had a consensus to rename the page to COVID-19 pandemic, would anyone know how to fix this? Thanks. TheGreatSG'rean (talk) 06:46, 4 May 2020 (UTC)[reply]

Inconsistency between the total number of confirmed cases and the addition of all confirmed cases for weeks

[edit]

Please, see Template talk:COVID-19 pandemic data#Inconsistency between the total number of confirmed cases and the addition of all confirmed cases for weeks. Lofhi (talk) 13:09, 23 May 2020 (UTC)[reply]

Important information

[edit]

Hi there. Ive been reading the Covid pandemic page. The Covid death toll reached 3 million many days ago. However, for the last few days this article was displaying a figure less than 3 million. For example, 2.9 or 2.89 million. Feel free to look at previous versions. Any reason for this? P.S. I'm not saying that the editors are always misleading or disruptive. 123.103.210.114 (talk) 06:22, 22 April 2021 (UTC)[reply]

No. of recoveries

[edit]

@JohnFromPinckney: The JHU source appears to have stopped publishing recoveries and territories data, replacing them with vaccination data. H78c67c (talk) 16:14, 15 May 2021 (UTC)[reply]